Hu et al. focused on somatic copy number alterations (SCNAs) of lncRNAs. By analyzing a genome-wide survey on SCNAs of lncRNA in 2394 tumor specimens from 12 types of cancer, they identified a correlation between copy number of focally amplified lncRNA on chromosome 1 (FAL1) and progression of ovarian cancer. They found out that FAL1 interacts and stabilizes the epigenetic repressor BMI1 (the PRC1 core protein) so as to repress p21 expression. Silencing FAL1 with siRNAs significantly blocked tumor growth in vivo. The interaction between FAL1 and BMI1 highlights how lncRNAs can take a central role in epigenetic regulation of p21 (80, 81).

BRAF-activated non-coding RNA (BANCR) is a lncRNA required for cell migration in different types of cancer (82–84). Shi et al. investigated its contribution to colorectal cancer. They showed that BANCR expression was significantly decreased in colorectal cancer, and when BANCR is overexpressed, it slowed down the growth of colorectal cancer. They also suggested that pcDNA-BANCR-mediated proliferation was linked to activation of G0/G1 cell cycle arrest through regulation of p21. Effects of BANCR were probably post-transcriptional since protein expression levels of p21 increased while that of mRNA remained unchanged. In brief, this article suggests that downregulation of BANCR contributes to the translational regulation of p21 protein (85).

LincRNA-p21 plays a role in regulating p21 levels. Upon DNA damage, p53 activates lincRNA-p21 that is located next to the p21 gene. In the nucleus, lincRNA-p21 activates transcription of p21 by recruiting hnRNP-K to the promoter region of p21 (86). Using conditional lincRNA-p21 knockout mouse model, Dimitrova et al. showed that loss of lincRNA-p21 could block p21 expression and it can activate mouse embryonic fibroblasts (MEFs) expansion. In the cytoplasm, lincRNA-p21 bound HuR recruits the let-7/Ago2 complex to destabilize lincRNA-p21. LincRNA-p21 also associates with its target mRNAs by base pairing and suppresses their translation with the Rck RNA helicase (86, 87).

The transcript of metastasis-associated lung adenocarcinoma transcript 1 (MALAT1) gene was first identified in 1997 (88), and it is highly expressed in several type of cancers (89). In 2010, MALAT1 was reported as a highly abundant nuclear transcript localized to the nuclear speckles (a region which is enriched in pre-mRNA splicing factors) (90, 91). In 2013, Tripathi et al. performed genome-wide transcriptome analyses in human fibroblasts. Their result indicated that MALAT1 controlled the expression of cell cycle genes and was required for cell cycle progression (92). They also showed that depletion of MALAT1 leads to activation of the p53 pathway and p53 is required for MALAT1 dependent cell cycle arrest. Moreover, MALAT1-depleted cells have low levels of B-MYB (Mybl2, an oncogenic transcription factor), leading to the discovery of aberrant alternative splicing of B-MYB pre-mRNA. However, the exact mechanism of how MALAT1 impacts p53 signaling needs further investigation (91, 92).

In 1991, Sakamoto et al. cloned 7SL sequence partially (93). They observed the 7SL would block proliferation of HeLa cells. Recently, Abdelmohsen et al. showed that 7SL forms a partial hybrid with the 3′-untranslated region (UTR) of p53 mRNA. The association of 7SL with p53 mRNA reduced p53 translation. When 7SL were silenced, it increased the interaction between HuR and TP53 mRNA, stabilizing TP53 mRNA and increasing p53 translation. The antagonist activity between 7SL and HuR for TP53 3′ UTR contributes to the fine tuning of p53 translation (94).

Lazorthes et al. have analyzed the differentially expressed strand-specific transcription during oncogene-induced senescence in human. They identified that most vlincRNAs (>50 kb) are induced during senescence. They showed that VAD, one of the antisense vlincRNAs, was highly expressed during senescence and it is essential for maintaining senescence phenotype. VAD regulates chromatin structure in cis and activates gene expression in trans at the INK4 locus. Moreover, VAD blocks the recruitment of H2A.Z, repressive histone variant, into the proximity of INK4 gene in senescent cells and maintains transcription of it (95).

Antisense non-coding RNA in the INK4 locus is produced from 9p21 on human chromosome locus where p16^INK4A and p15^INK4B (p15 in mouse) are transcribed. SNPs in ANRIL gene are linked to various diseases in genome-wide association studies (GWAS). Additionally, silencing of this locus affects cellular proliferation, and it has been found that polycomb complex is recruited to this region by ANRIL via SUZ12 (a component of polycomb complex). Silencing of ANRIL produces senescent cells that are stained positive for beta-galactosidase. Current research suggests that ANRIL recruits PRC1 and PRC2 to INK4 region to initiate and maintain silenced chromatin signature: H3K27ME and H2AK119ub1 (96, 97). Therefore, ANRIL maintains the proliferative state by blocking senescence genes.

MIR31 host gene is located 400 kb upstream of the p16^INK4A locus on chromosome 9 in human. Previously, decreased expression of MIR31HG has been shown to reduce cell growth and activate strong senescence phenotype (98). Recently, Montes et al. found that the lncRNA MIR31HG was highly expressed in senescence and silencing of MIR31HG activates p16^INK4A expression (99). In pre-senescent cells, MIR31HG is located in both nucleus and cytoplasm, but after B-RAF (proto-oncogene) transcription, MIR31HG is specifically present in the cytoplasm. Their results showed that MIR31HG could bind to both p16^INK4A and MIR31HG genomic regions. Additionally, polycomb group (PcG) proteins and MIR31HG are necessary for PcG-mediated repression of the p16^INK4A locus. Hence, their data suggested a new lncRNA-mediated inhibition of the p16^INK4A (99).

Kumar et al. discovered that CAPERα/TBX3 repressor complex was required to prevent senescence in primary cells and mouse embryos. CAPERα/TBX3 complex regulates chromatin structure and represses transcription of CDKN2A-p16. Inhibiting CAPERα/TBX3 induced the lncRNA UCA1 expression, in which UCA1 overexpression induced senescence. In proliferating cells, hnRNPA1 can bind and destabilize p16^INK4A mRNA. However, during senescence, UCA1 expression stabilizes p16^INK4A mRNAs by disrupting the association of hnRNP A1 with p16^INK4A mRNA (100, 101).

Abdelmohsen et al. identified differentially expressed novel lncRNAs between young versus old human diploid WI-38 fibroblasts populations. These populations are 32 days apart from each other in doubling time. Among those novel lncRNAs, they selected three lncRNAs to further study their involvement in cellular senescence and proliferation. Decreasing expression of those lncRNAs resulted in increased p53 expression, thereby affecting cell survival and senescence (102). However, exact mechanism requires further study.

Using human fibroblast cells, Wu et al. performed genome-wide screening of lncRNA expression in cellular senescence. They have identified lncRNAs with differential expressions in senescent cells in contrast with young cells. Within these lncRNAs, they further selected a specific senescence-associated lncRNA (SALNR) that has low expression in senescent cells. When they overexpressed the SALNR, cellular senescence was delayed.

Furthermore, they found that SALNR physically interacts with NF90 (nuclear factor of activated T-cells, 90 kDa), which is able to suppress biogenesis of senescence-associated miRNAs, such as miR-22 and miR-181a. Inhibiting NF90 resulted in premature senescence since NF90 is the suppressor of senescence miRNAs.

Moreover, when cells were exposed to Ras-induced stress (activates senescence), NF90 gets translocated to nucleolus and NF90 can no longer suppress senescence-associated miRNA biogenesis, which can be rescued by SALNR overexpression. Thus, SALNR antagonizes NF90 translocation into nucleolus and rescues its inhibitory activity on senescence-associated miRNA expression. Their data suggest that lncRNA SALNR controls cellular senescence through modulating NF90 localization (103).

First, telomeric repeat-containing RNA has been identified in yeast and it has a specific hotspot in the subject of senescence. It forms DNA–RNA pairing with telomere region at R-loop formation that contains G-quadruplex, TERRA, and TERRA-binding proteins. TERRA also increases upon DNA damage at telomere region to involve in recombination and DNA repair. Arnoult et al. showed that TERRA expression decreases upon telomere elongation in multiple cancer cell lines (104). Elongated telomeres have less TERRA expression due to tightly packed telomeric repeats with HP1α and H3K9me3, which make chromosome inaccessible to the transcription machinery.

In parallel, increased levels of TERRA was also found in several primary human tumors containing short telomeres (105). In other words, TERRA is reversely correlated with telomere length.

On a different research, TERRA was shown to maintain the telomere length in the lack of telomerase enzyme as explained above. RNA–DNA complex formed between TERRA and telomeres postpones the senescence by activating homologous recombination at telomeres (106). These suggest multiple functions for TERRA on telomere length and maintenance; however, further research is required to unravel the detailed mechanism of how TERRA functions in telomere and telomerase during aging (107–109).

Villegas et al. have first reported a mitochondrial transcript of 2374 nucleotides whose expression correlates with the cellular replication (110). Following up this project, Burzio et al. have reported the existence of two antisense mitochondrial transcripts that were expressed in proliferating cells from healthy human tissues but decreased in tumor samples (111). Recently, Bianchessi et al. showed that two miRNAs, hsa-miR-4485 and hsa-miR-1973, have perfect homology with ASncmtRNA-2. Both miRNAs are upregulated in endothelial cells at replicative senescence. ASncmtRNA-2 overexpression results in an increased expression of hsa-miR-4485. It induces cell cycle arrest at G2/M phase. Moreover, the overexpression of the two miRNAs shows the similar phenotype, cell cycle delay in both the G1 and G2/M phases. They proposed that ASncmtRNA-2 is involved in replicative senescence in endothelial cells by causing cell cycle arrest in the transition from G2/M to G1 (112).

The human lncRNA HOX transcript antisense RNA (HOTAIR) is transcribed from the HOXC locus of homeobox genes as an approximately 2.2-kb RNA (113). It binds to the enhancer of zeste homolog 2 (EZH2) in the polycomb repressive complex 2 (PRC2) and to lysine-specific demethylase 1 (LSD1), which are chromatin remodeling factors (113, 114). Both complexes lead to gene silencing and are recruited in trans to hundreds of genomic sites by HOTAIR. Recently, Ozes et al demonstrated a novel function of HOTAIR in DDR, which is regulating the expression of genes to block proliferation. Their results showed that after induction of DNA damage, HOTAIR activates the cellular senescence through NF-κB pathway. Furthermore, they demonstrated the presence of a positive-feedback loop that induces HOTAIR. This activates DDR and sustains NF-κB expression (115). Recently, Yoon et al. reported a function of lncRNA HOTAIR as a scaffold for ubiquitin-mediated proteolysis in posttranslational control (116). First, HOTAIR binds to E3 ubiquitin ligases through RNA-binding domains, Dzip3 and Mex3b, and to their ubiquitination targets: Ataxin-1 and Snurportin-1, respectively. As a result, HOTAIR assists the ubiquitination of Ataxin-1 by Dzip3 and Snurportin-1 by Mex3b and speeds up their degradation. HOTAIR expression is increased in senescent cells, resulting in the degradation of Ataxin-1 and Snurportin-1 (116). Their results suggested a novel scaffold function for a lncRNA that can impact on senescence.

With the help of super high-resolution tiling array analysis in CDKN1 region, Puvvula et al. identified P21-associated ncRNA DNA damage activated (PANDA) as a lncRNA, which is induced upon DNA damage.

P21-associated ncRNA DNA damage activated plays a dual role in proliferating cells versus senescent cells. In proliferating cells, scaffold-attachment-factor A/hnRNPU (SAFA) interacts with lncRNA PANDA to block senescence genes by recruiting PRC1 and PRC2 to senescence activating genes loci including CDKN1. Silencing SAFA and PANDA induces cell cycle arrest, and it also causes senescence in proliferating cells by allowing transcription of senescence activating genes.

In senescent cells, SAFA–PANDA–PRC interactions are blocked therefore, pro-senescence genes are transcribed. PANDA inhibits the expression of proliferative genes by sequestering the transcription factor NF-YA from occupying its target promoters in senescent cells. Depletion of PANDA in the senescent cells reactivates proliferative genes and allows cells to replicate (117).